L’objectif du travail est d’évaluer le niveau des contaminants (éléments à l’état de traces et pesticides) dans quatre types de produits alimentaires (pain, fromage, végétaux et fruits) et eau potable considérés parmi les produits les plus consommés par les Libanais suivant l’étude du panier de la ménagère faite en 2006. Une méthode analytique pour l’analyse multi élémentaire a été optimisée pour augmenter la sensibilité et réduire les interférences polyatomiques en utilisant la technique de spectrométrie de masse de haute résolution par plasma induit pour la détermination de 20 éléments mineurs et à l’état de trace afin d’évaluer la valeur nutritive et le niveau de contamination. Pour l’analyse des résidus de pesticides, une méthode analytique a été développé pour la détermination des résidus dans des fromages riches en matières grasses (2-25%). La détermination des résidus de pesticides dans le pain, les végétaux et les fruits a été accomplie suivant la méthode QuEChERS. La surveillance du taux des contaminants a été faite sur des échantillons collectés dans des marches de cinq régions géographiques au Liban durant les saisons humide et sèches. / The objective of this work was to evaluate the level of food contaminants, trace elements and pesticides residues, in four types of commodities bread, white cheese, vegetables and fruits, and drinking water considered as the most consumed by the Lebanese population based on the Food Basket Study on 2006. A multielemental method was optimized in terms of sensitivity and inference elimination by using high resolution sector field inductively coupled plasma mass spectrometry (HR ICP/MS) for the determination of 20 minor and trace elements in the selected foods to assess the nutritional contribution and the level of contamination. For pesticides residues analysis, an analytical method was developed for the quantification of residues in cheese (2-25% fat). Bread, vegetables and fruit were analysed according to the QuEChERS method. For the monitoring study, the samples were collected during two seasons, wet and dry, from five geographical regions in Lebanon.

O objetivo principal deste estudo foi o desenvolvimento e validação de metodologia de análise de multiresíduos para quantificação de resíduos de pesticidas em manga. Foram analisados 32 pesticidas, sendo dos seguintes grupos: 9 Organofosforados, 17 Organoclorados e 6 triazois sendo que neste mesmo método foi analisado novamente o Organofosforado fention. Os organofosforados foram analisados em cromatógrafo a gás (CG) acoplado ao detector fotométrico de chamas e os organoclorados foram analisados em cromatógrafo a gás (CG) acoplado ao detector de captura de elétrons (ECD). Os triazois foram analisados em cromatógrafo a gás acoplado a espectrômetro de massa (CG/MS). A curva de linearidade para todos os padrões analíticos, foi realizada em triplicata e com 6 diferentes concentrações. As fortificações das amostras foram realizadas em dois níveis de concentração, no limite do método e 10 vezes o limite do método, sendo que a concentração do limite de quantificação do método (LOQ) foi realizada com sete repetições (N= 7) e a concentração de 10 vezes o LOQ foi realizada com N= 5, usando-se mangas tratadas no sistema orgânico (livres de pesticidas). Os métodos desenvolvidos e validados neste estudo apresentaram-se: a) Linear na faixa de: 0,1 - 4,0 µg/mL para o triclorfon e malation; 0,05 – 2,0 µg/mL para os demais organofosforados; 0,05 – 1,6 µg/mL para os organoclorados; 0,1 – 1,0 µg/mL para o paclobutrazol; 0,25 - 2,5 µg/mL para o fention e bromuconazol; 0,5 – 5,0 µg/mL para o tetraconazol e tebuconazol e 1,0 – 10 µg/mL para o procloraz e difenoconazol respectivamente, uma vez que o coeficiente de determinação foi maior que 0,99 para todos os ativos. b) Específicos, pois o nível de contaminação e/ou interferente na análise do branco dos reagentes e da amostra testemunha, foram inferiores a 30% do limite de quantificação. c) Exato, Preciso e Repetitivo, pois todos os ativos apresentaram recuperação dentro do intervalo de aceitabilidade de 70 a 120%, com coeficiente de variação porcentual (CV%) inferior a 15% em relação à média global de todos os níveis de fortificação. Foram também analisados os mesmos pesticidas em 15 amostras de mangas compradas nos supermercados de Piracicaba, procedentes do Nordeste do Brasil. Em nenhuma das amostras foram encontrados resíduos acima do limite estabelecido por lei ou acima do limite de quantificação deste método, no caso dos pesticidas não autorizados. / The principal objective of this study was the development and validation of a multiresidue analytical methodology for the quantification of pesticide residues in mangos. Quantitative analyses were made for 32 pesticides, from the following classes: organophosphorus (9), organochlorine (17) and triazole (6). The results from the method include the first analysis for the organophosphorus pesticide fenthion. The organophosphorus pesticides were analyzed by a gas chromatograph (GC) coupled to a flame photometric detector, while the organochlorine pesticides were analyzed using a gas chromatograph (GC) coupled to an electron capture detector (ECD). The triazoles were analyzed with a gas chromatograph coupled to a mass spectrometer (GC/MS). For each of the analytes, calibration curves were constructed in triplicate using solutions with six different concentrations, prepared from the analytical standards. Recovery experiments were performed at two levels of fortification, the limit of quantification of the method (LOQ) and 10 times this limit. Seven repetitions (N = 7) were performed at the limit of quantification, and five (N = 5) at the 10 times higher level of fortification. Mangos raised by organic farming methods (free from pesticides) were used for the recovery experiments. The methods developed and validated in this study have the following characteristics: a) linear range: 0.1 – 4.0 µg/mL for trichlorfon and malathion; 0.05 – 2.0 µg/mL for the other organophosphorus pesticides; 0.05 – 1.6 µg/mL for the organochloro pesticides; 0.1 – 1.0 µg/mL for paclobutrazol; 0.25 – 2.5 µg/mL for fenthion and bromuconazole; 0.5 – 5.0 µg/mL for tetraconazole and tebuconazole; and 1.0 – 10.0 µg/mL for prochloraz and difenoconazole, in each case the coefficient of determination was greater than 0.99. b) Specific: thus the level of contamination and/or intereference in analysis from the reagent blank and the control sample amounted to less than 30% of the limit of quantification. c) Exact, Precise and Repeatable, for each of the analytes the recovery was found to be within the acceptable interval of 70 to 120%, and with a coefficient of variation, expressed as a percentage, of less than 15% in relation to the global average of the results obtained at the two levels of fortification. Analyses were made for the same pesticides in 15 samples of mangos bought from supermarkets in Piracicaba, which were raised in the Northeast of Brazil. None of these samples contained residue concentrations above the limit established by law or, in the case of non-authorized pesticides, above the limit of the quantification of the current methodology.